Relay true bypass power Consumption

[GibsonGM]

I should've brought that up earlier, R.G...I NEVER will use a relay for this purpose unless it's a DPDT latching variety with TWO coils.  SO much easier to deal with!!!  I found a type I like and bought about 20 of them, all set.  They work on a pulse, very easy to control.   Use anywhere...pedals, amps....

[jackwithoneye]

I finally found time to get back to work on this circuit, and breadboarded all the relay switching circuit from scratch.

Quick recap :
-FX circuit is a Tone bender like with npn transistors
-Switching circuit : 9V Relay (ry9wk, non-latching, 540ohm coil)+ ne555

Power consumption is now ok with this 9V non-latching relay (for FX+switching : more or less 39mA when switched on, and 10mA when off) and i like the fact that when the circuit/pedal is not powered, it comes back to true bypass

ISSUES :
-popping/ticking when switching when there's no audio
-bouncing/triggering


TESTS :
- I ran some tests with seperated/independant power supplies for the circuits, and it still popping when there is no audio volume. I'd seperated "audio ground" and "switching ground" as long as i could on the breadboard, until they join at the power connector.  Confirming that it could come from coupling inside the relay itself or power glitches, i guess?

- i'd improve the power supply part with Rob's recommandation : +9V ---diode ---- 100uF to ground [this rail to switching ckt pwr] ---> 100R ---> 100uF to ground [this rail to effect power]
didn't improve popping

- I have a bouncing problem on the triggering (about 5-10% of the time, was a lot less when it was on pcb), tried to find the right R1/C1 combination, but 220K/220nf remain the most reliable, but not enough....

-I also tried to put some 100nF grounded cap in different places (coil, transistor base) but still, no real improvement.

I'm a little bit lost to be honest. I hesitate to startover with new schematic which would be easier to manage
or go back to 3pdt but , i really like the soft touch footswich feeling :
After spending more than 100$ in relays and ne555 and stuffs, i would appreciate to get it working
Dual coil latching relays are pretty hard to find and expensive where i live in europe. (about 6-7$ to get one).
I have no skills at all in coding a PIC (to substitute ne555) to command a relay

[Rob Strand]

Upfront, I want to be very clear that the ground on the relay contact (relay pin 6) should *not* connect to the ground on Q1's emitter/C1/IC pin1.   Such a connection will make it hard to separate the possible causes in the list I wrote in the summary.  In fact with that connection you might not ever be able to solve the click problem.  You should think of relay pin 6 as belonging to the audio/effect circuit.

It would be very helpful if you can spell-out what you had here in your tests.  If relay pin 6 was on the switching circuit ground then the best thing to do would be to redo the tests with relay pin 6 connected to the audio ground.

Whatever the test configuration relay pin 6 should always be connected to the audio circuit.

TESTS :
- I ran some tests with seperated/independant power supplies for the circuits, and it still popping when there is no audio volume. I'd seperated "audio ground" and "switching ground" as long as i could on the breadboard, until they join at the power connector.  Confirming that it could come from coupling inside the relay itself or power glitches, i guess?

Excluding my opening caveat that would be the conclusion.   However when you say,

I'd seperated "audio ground" and "switching ground" as long as i could on the breadboard, until they join at the power connector.

It makes me think the ground wiring isn't correct.

The starting point in the test is a configuration where no grounds between the two parts of the circuit (other that relay pin 6).  After that you start changing the setup.

- i'd improve the power supply part with Rob's recommandation : +9V ---diode ---- 100uF to ground [this rail to switching ckt pwr] ---> 100R ---> 100uF to ground [this rail to effect power]
didn't improve popping

Using separate power supplies *should* make that filtering unnecessary.  However, in the final solution (when we get there) it is almost certain a filter (or filters) will be required.   The stuff RG posted is along the same lines.

The reason is the changes in current from the relay will cause voltage the *incoming* power supply to dip.   No matter how perfect you do the power wiring and ground wiring you can't stop that dipping.  So instead we must add filtering to the supply to hide it.

- I have a bouncing problem on the triggering (about 5-10% of the time, was a lot less when it was on pcb), tried to find the right R1/C1 combination, but 220K/220nf remain the most reliable, but not enough....

I'm not convinced the time constants in the circuit are long enough to debounce the switch.    You can try increasing R4 and R5.   Try a very large value like 220k or 470k for each.
It needs a closer look.  The R4 and R5 changes aren't it.

OK, pin 4 is open.  Connect it to pin 8.

-I also tried to put some 100nF grounded cap in different places (coil, transistor base) but still, no real improvement.

At this point there might be bigger issues.

Editted

[jackwithoneye]

Upfront, I want to be very clear that the ground on the relay contact (relay pin 6) should *not* connect to the ground on Q1's emitter/C1/IC pin1.   Such a connection will make it hard to separate the possible causes in the list I wrote in the summary.  In fact with that connection you might not ever be able to solve the click problem.  You should think of relay pin 6 as belonging to the audio/effect circuit.

Yes, i've been doing this, it was clear in my mind that relay's pin 6 is part of the "audio circuit ground"

Running seperated power supplies (and fully seperated grounds on two differents breadboards) for FX and Switching circuits gives me a buzz/noisy in the audio circuit. However, i just put the FX volume at zero to have silent FX output and certify there's still the pop/tick when switching. (same when the FX part is not powered at all)

Joining grounds at the DC/Jacks grounds make the Audio path clean, but still popping when switching.

It makes me think that the pop/tick is coming wether the relay contact coil coupling or the current peak on the NE555 state change.


In the ideal PCB trace routing ground for pcb design, correct me if i misunderstood, because i might not get right the concept (kind of star grounding?):
-I would use a ground plane on the pcb for the audio Path ground, including to an "Audio Ground pad", which would not extend to the part where the relay coil is, to avoid coupling.
-I would make a "Switch ground" trace joining all the switching circuit grounds (except Relay pin 6!) which would lead to a "switch ground" pad
-Connect the "switch ground pad" and the "audio ground pad" to the DC connector ground.
-Having In&Out jack sleeves connectors directly connected to the audio ground plane through dedicated pads and not to the DC connector.

[anotherjim]

I would filter the audio circuit and relay switching power supplies separately.

R1 is chosen as 1/10th of the relay coil resistance. It doesn't implement reduced hold current so it doesn't affect the 555 timer latch. For reduced old current, another RC network can be inserted into the feed to the relay coil after C2.

[R.G.]

A.J. is right - the filtering you need for the relay coil and switch transistor is different from what you need for the rest of the circuit.

With relays, it's possible to do all the usual things wrong and get clicks and pops. The relay adds two new ones. These are inducing power line and ground transients to the analog circuit power line with sudden current needs, and capacitive coupling of the (large) voltage change across the coil into the audio signal line you're running through the relay.
Let's do capacitive first. It's easiest to understand, but maybe hardest to do anything about. You want your audio signal to be without clicks and pops when there's no audio signal through it and the amps are turned up to ear-bleed volume. The wires the signal run on have impedances at each end - a 1M-ish one at the pedal parallel and the pickup at the other. We'll model the pickup as a 1H to 2H inductor paralleled by 4K to 20K of wire resistance. The fact that it's inductive (and distributed-capacitive) mean it has a high-ish impedance at the high end of audio, and that all is paralleled to the summed-up cable capacitance. That is a modestly complicated network, but we can ignore the cable capacitance and think only of the inductance sending a signal voltage to the 1M input. This is a gross oversimplification, but it's useful to understand. That cable also has a capacitance to ... everything that's conductive in the entire universe. That capacitance trails off as the inverse of the square of the distance from the signal metal to [whatever else]. It's an unwanted input capacitance.
Any voltage change on the outside of that stray capacitor gets coupled through as (oversimplifying again) a high passed signal through the small stray capacitor and the 1M/1-2H/cable capacitance/etc. In any real setup, there's a lot of attenuation, but the attenuation gets smaller as the frequency gets higher.
Run an input signal through a relay, and you're putting part of your signal wire very, very close to a coil of wire that you are going to whack with 5V to 9V of coil voltage, then release. The turn-on voltage appears full voltage, then sags for micro- or milli-seconds as the coil voltage rises. At turn off, the voltage rises semi-instantaneously from just the resistive voltage across the coil to the power supply plus the flyback voltage of the relay coil. Very fast signal edge, couples well across the stray capacitance to the signal wires in the relay.
Without a shielded relay, you're not every going to make the coupled signal go to zero. You can make it smaller until it's not noticeable by slowing down the voltage rise and fall. Ideally, you'd try some kind of current ramping up/down on the relay coil. I've tried that. It works, but it's complicated. The one-transistor integrator from the geofex circuit work well enough to suppress coupled relay coil ticks down to the units of millivolts. That turned out to be not objectionable to me, my boss, or our customers, so we went with it. You have to turn the amp up and put your ear near the speaker to hear this in most setups. A theoretically better solution would be to make the signal line impedance be much lower, as this would make the stray capacitor have to move a lot more current to get the signal voltage to change, but high impedance signal lines on inputs is what we have to have for other reasons. So lower signal impedance levels isn't really a practical answer.
Power supply transients from turning the relay coil on and off can be a little baffling. After all non one has an isolated power supply in a pedal, do they? So the relay has to share ground somewhere, even if it's just star grounded back to the power supply.
What has worked for me is a variant of what I was describing before and what A.J. was getting at - isolating the relay coil current loop.
We don't much care, within some limits, what the DC conditions are. We just want sudden current edges, both on and off, to not cause +9V (or 12, or whatever) to tick up and down, and we really don't want those same spikes to raise and lower ground, as that sends a <tick> directly into our circuit inputs. We can "isolate" our relay ground from analog/signal ground by making the relay coil current really, really want to stay in a nice tight loop from a capacitor near the coil, through the coil, and back to the capacitor - and not back to audio ground.
Take a resistor from(for instance) +9V to the + end of a cap, ground the other end of the cap. If we hook the relay coil and series switch transistor across that cap, then turning the transistor on and off operates the relay. The cap charges up to 9V when the relay is off, provides the current for the coil when the transistor is turned on, and does not suck transients from the power supply. The charge stored in the cap >wants< to go back to that cap, not the shared power supply.
When you turn the transistor on, cap voltage appears across the coil. The coil current ramps up at a rate governed by the inductor equation, V = L * di/dt. At the time just after the switch turns on, the cap voltage is still 9V, the coil voltage is 9V, but the current has not ramped up from zero. As the milliseconds plod on, the inductor current ramps up at di/dt = V/L, and the capacitor voltage, governed by I = C* dV/dt ramps down. If the cap was not there, the current would have to come through the power supply wire resistance and return through the ground path to the power supply.
With the cap there, the current into the coil is ramped up, the cap voltage is ramped down, and the difference between the analog power 9V and the capacitor voltage increases through the series resistor to the cap. So a slowly increasing current from the analog power supply voltage into the resistor/cap/relay circuit happens, and juggling the resistor and the cap you can make this voltage change on the analog power supply as slow as you want. What you want is to have this voltage change to have a frequency content below audio, or below the pass band of your pedal and amp. If that's true, you will not hear a transient.
What about the ground current? Won't it make a tick? Maybe. You can often get away with a gimmick, a low-ohms resistor in series between the capacitor negative and the main circuit ground. I've only had to do this once over the years. Setting up a nice, tight, everything-side-by-side circuit for the relay coil, switch transistor, decoupling cap and resistor(s) can often be magic. But it still won't make the transient be zero - just unnoticeable.

Well, I see I just set off typing again. Ask questions about what is muddy.

Oh, and get rid of that 555. Just do it. Use a CMOS 7555 equivalent. Just do it. Some 555 circuits can never be quieted, and I've tried.

[jackwithoneye]

wow, just wow
Thanks for your time, that's wonderfull to have all you guys helping me with that kind of precious knowledge.

I have to re-read all this few(thousands) times, to get back with not "too" dummy questions, because it's way above my skills. I want to learn, so i'm to trying to understand the most i can. Because right now, i just dont know how to integrate this and link it with my circuit. I'll try AJ's power setup asap
I guess that the fact that everything is breadboarded doesnt help with capacitance management.....

I'll get rid of the ne555, that i get. I'll buy a bunch of the cmos version ICM7555IPAZ. (that's the easy part )
datasheet : https://www.renesas.com/us/en/document/dst/icm7555-icm7556-datasheet

[Rob Strand]

Yes, i've been doing this, it was clear in my mind that relay's pin 6 is part of the "audio circuit ground"

OK, great.

Running seperated power supplies (and fully seperated grounds on two differents breadboards) for FX and Switching circuits gives me a buzz/noisy in the audio circuit. However, i just put the FX volume at zero to have silent FX output and certify there's still the pop/tick when switching. (same when the FX part is not powered at all)

Joining grounds at the DC/Jacks grounds make the Audio path clean, but still popping when switching.

The buzz problem is a different issue.   It's common on breadboards.  There's plenty of ways to reduce it.  For example, shielded wires on each of the relay contacts then connect the both shields to audio ground and only one of the other ends to relay pin 6.   You can also sit both breadboards on a metal sheet (aluminum foil works) connected to audio ground.  Some breadboards have an aluminum backing already - I ground this more often than not.   You can experiment with these if you like as it's common issue with breadboard.  However, it's more for your own interest and it's probably not to help your click (it could help a small amount).

It makes me think that the pop/tick is coming wether the relay contact coil coupling or the current peak on the NE555 state change.

I don't think it's the NE555 but to *prove* it you can wire a switch to the base transistor resistor, the switch supply rail and ground.  In effect replacing the NE555 output with a switch.

You should also try another experiment where you replace the relay contacts with a switch.   The reason is it proves your current set-up can work at all.  Maybe there's a problem elsewhere.  This test will give some peace of mind.   It would be really annoying if there's some other bug contributing to the click.

For tricky problems it is a good policy to take a step backwards and *prove* what you think is working actually does work.

In the ideal PCB trace routing ground for pcb design, correct me if i misunderstood, because i might not get right the concept (kind of star grounding?):
-I would use a ground plane on the pcb for the audio Path ground, including to an "Audio Ground pad", which would not extend to the part where the relay coil is, to avoid coupling.
-I would make a "Switch ground" trace joining all the switching circuit grounds (except Relay pin 6!) which would lead to a "switch ground" pad
-Connect the "switch ground pad" and the "audio ground pad" to the DC connector ground.
-Having In&Out jack sleeves connectors directly connected to the audio ground plane through dedicated pads and not to the DC connector.

That would all be for doing the best you can but if it's going to work then you should be able to follow the diagram anotherjim has kindly drawn up.   The physical way the ground and power connections are made should follow anotherjims drawing.   For experiment you could add another 100uF cap from where R1 and R2 join back to the common DC input ground at the top left.   You could also try powering the LED from the from the other side of R1.

That should be enough to fix any power and ground issues.   Your final circuit should keep that configuration as anything less is asking for trouble.

There is still one crack that can cause clicking and that's coupling inside the relay.   That's quite possible and I'm thinking even likely.  By checking and doing all the above it's essentially proving that's the only cause left to blame.

The problem is coming up with a solution.   We can't modify the relay or ground the internals.   The obvious solution is to slow down the signal of the collector of the relay transistor.    That's kind of where we go upto very early on but we didn't have any proof that coupling through the relay was the only cause.    That means adding resistors and capacitors.

Much more sinister solutions involve grounding the relay - terminal and switching the relay + terminal.   That would work only for particular relays.  The idea is a *particular model* of relay may have more coupling from the - terminal to contacts than the + terminal.  So by switching the + terminal it reduces the coupling.

It could be worse,  the same, or only a little bit better than before. You can only try it and live in hope. 

You would need to change the relay drive circuit to so that.  At this point I won't clutter the post with details.  However as an experiment it would be easier to switch the relay off and on with a switch (as per above experiment).

Slowing down the signal on the collector is kind of forcing success since it has to at least reduce the problem.

[R.G.]

The NE555 was a great chip for many things, and revolutionized timer chip use when it came out. But that was at a time when a few pesky current spikes on the power supply didn't matter much. The bipolar 555 can draw >an amp or more< directly from power to ground at the instant it switches.  It's kind of a current-mode Molotov %^&*tail.
This might be your problem, or, since there is almost never only one problem, it might only be an accomplice to the crime.

As Rob notes, using a pair of wires to connect/disconnect the relay coil would eliminate the 555's contribution. It's a quick, cheap test. Your description of the click as being very trebly, tick-not-pop, weighs in favor of the stray capacitance inside the relay, again as Rob notes. You could also (sneakily...) disconnect just the input and then just the output and switch it to isolate whether it's from the input signal switching (I found this was more frequently the issue) or the output switching. Or both, equally.

Be very sure that the ticking is not an artifact of a small DC level difference on the output of the pedal, input of the amp, and maybe try different amps. Yep, found that issue before too.

[PRR]

Not to help with jackwithoneye's problem, but for general info:

The April 1967 issue of Electronics World is all about relays. Even Frye's column.
https://worldradiohistory.com/Archive-Electronics-World/60s/1967/Electronics-World-1967-04.pdf
https://worldradiohistory.com/Electronic_World_Master_Page.htm

[amptramp]

I realize the schematic drawn by anotherjim is somewhat simplified, but there has to be a base resistor from the timer to the transistor.  An R-C network or better yet, an R-C-R network from the timer to the transistor with the C grounded may slow the current rise and fall enough to get rid of some of the problem.

[antonis]

an R-C-R network

I was interpreting last R as a typo 'till realize you refer on it as cap shunt discharge resistor, Ron..

[anotherjim]

I wasn't interested in the 555 circuit so I didn't show any of it, but I had to show the BJT as it's attached to the relay. The base drive is inside the "Timer" block, natch.
Thing is, the audio circuits are discrete class-A so prone to passing supply rail noise via bias networks into the audio path & amplifying it - so, unless battery powered should have an independently filtered DC supply -  relay or no relay.

[jackwithoneye]

I'd guessed that Jim's schematic was focused on the power supply & filtering part, with the "timer circuit" as it was set to the bjt. I haven't any results to give yet, cause i fried some part doing other tests due to breadboard glitches. but i will get back to it. 

i draw this, based on the Jim's power and grounding supplies drawing


How do i choose the R-C-R(R2) values (pink square on the schematic) for slowing current rise?

Correct me if i'm wrong, as far as i understand, i have to do a seperated trace for VD (switching circuit ground) which lead to a pcb pad, connected to the DC connector ground (as for the Audio ground) right?
I'm using Eagle, and it doesn'nt recognize any difference, as GND and VD is the same net, so i have to remove the Audio ground plane on the pcb design where the switching circuit is, to be able to route VD to a seperated pad without ratnest messing this routing.

[R.G.]

Correct me if i'm wrong, as far as i understand, i have to do a seperated trace for VD (switching circuit ground) which lead to a pcb pad, connected to the DC connector ground (as for the Audio ground) right?

I would if I were you.
But I would suggest a BFC* as close to the relay as you can get it, and the switch transistor bridging the gap between the negative side of the relay coil and the negative side of the BFC. Any catch diodes go from the collector of the switch transistor to the positive of the BFC. This circuit lump needs to be as close together and compact, with the shortest possible traces connecting pieces. Don't go all loopy on the traces. A resistor to the general power supply, ideally leading back to the positive pin for power coming onto the PCB and no.where.else. would be a vast help.

*Big Freaking Capacitor

I'm using Eagle, and it doesn'nt recognize any difference, as GND and VD is the same net, so i have to remove the Audio ground plane on the pcb design where the switching circuit is, to be able to route VD to a seperated pad without ratnest messing this routing.

I don't use eagle, but I have run into this before. Do this: define a dummy component "resistor.ground.isolation" and put it between the general analog ground and the negative side of that BFC. Label the negative side of the BFC something poetic, like "relay ground". The other side of the "resistor" is general power ground, but you will route the trace from the resistor to the board negative power input pad without touching anything along the way.

"resistor.ground.isolation" can be any value, from zero (soldered in wire) to many megs. If you never, ever plan to put a resistor there, you can at the very end of layout when the traces are all done remove the resistor and hard-connect the two "grounds". I would suggest leaving it in and experimenting with values from soldered wire up to 10 or 22 ohms.

But it still might be relay capacitance clicking...

[jackwithoneye]

I would filter the audio circuit and relay switching power supplies separately.

R1 is chosen as 1/10th of the relay coil resistance. It doesn't implement reduced hold current so it doesn't affect the 555 timer latch. For reduced old current, another RC network can be inserted into the feed to the relay coil after C2.

For info, the relay i use, RY9WK, has, as far i as i know a non-polarized coil, there no particular + or - coil pins on the datasheet
The timer is still NE555, as i haven't received the cmos version yet.

Test 1 : Powered the relay coil independantly, without the timer circuit, straight from an independant power supply+ momentary switch, without filtering
=>still got the tick+pop when switching

Test 2 : I applied strictly, on my breadboard, Jim's power supply filtering above's schematic for the switching&FX, fed from one power supply.
=> Good improvement, stability, buzzing is a lot better.
The switching tick/pop is definitly better, the "tick" is almost gone, but there's still a real pop going through the FX circuit (i can filter it with the high pass in the audio FX circuit) when switching ON, only.
Power consumption is all good with 24mA when on and 9mA when off.

test3 :
I had tried (maybe in a bad way, due to my misunderstanding? see below schematic) RG's BFC:
tried (in both direction) 100uF,47uF,10uF
It's actually slowing down the relay release (switching off), and making the "Zeeeew" light saber sound during this release, but doesn't affect the switching ON pop.



i did not implement yet any RCR (like amptramp said) to slow down current rise as i don't have a clue of the values needed, and it might still depend of cmos ne555 (ICM7555IPAZ) that i'll try, right?

And again, big thank you all for helping!

[Rob Strand]

Test 1 : Powered the relay coil independantly, without the timer circuit, straight from an independant power supply+ momentary switch, without filtering
=>still got the tick+pop when switching

Test 2 : I applied strictly, on my breadboard, Jim's power supply filtering above's schematic for the switching&FX, fed from one power supply.
=> Good improvement, stability, buzzing is a lot better.

Those results don't make sense as jim's circuit doesn't change the switching times.   It separates the power and test 1 already *is* separating the power.  In test 1 did you have a diode on the relay?

As a sanity test you should try replacing the relay entirely with a normal switch.  If that ticks/pops you have another issue which cannot be solved by playing with the relay circuit.

test3 :
I had tried (maybe in a bad way, due to my misunderstanding? see below schematic) RG's BFC:
tried (in both direction) 100uF,47uF,10uF
It's actually slowing down the relay release (switching off), and making the "Zeeeew" light saber sound during this release, but doesn't affect the switching ON pop.

That connection works better for switching off.  When switching on the transistor shorts out the cap so the on-time is fast.  The turn-on speed is limited only by the transistor's ability to pass current (set by the base resistor).   So basically this circuit has extreme asymmetry in the turn on and turn off times.

To filter both on and off you need  different configuration.   Unfortunately most simple circuits like that one will end-up with the turn on being somewhat faster than the turn on.   You can for example add a resistor in between the relay+diode+added cap and the collector of the transistor.   The larger the resistor the slower the turn on time.   However, due to voltage drops, the resistor can't be too larger otherwise the relay won't turn on correctly.   The other way to make the turn-on time longer is to increase the added cap.  However that will also increase the turn-off time.   This is where the asymmetry in the on and off times becomes hard to work around.


If you still have the 10k base resistor, it might be worth trying a cap across the base and collector of the transistor again.

[anotherjim]

Out on a breadboard, I think any switching contacts that are not screened from radiated noise might create clicks in the audio. This can even happen if the noise isn't audible. If there is any stray signal present there is a short time when if the contact voltage is non-zero it gets transferred between paths as stray capacitance holds it momentarily as some random DC level. Ordinary PCB relays don't usually have any screening of the contacts and they are sticking up inside the plastic case like aerials.

This is the general layout in a PCB relay. Differences might be the coil is rotated 90deg. The contact operating part should be electrically insulated from the armature otherwise the contact capacitance will change between states.
If the coil is in the middle, like the following illustration, it should probably not be used for audio as the common contact is more prone to noise pickup and capacitance changes between states.

[Rob Strand]

Out on a breadboard, I think any switching contacts that are not screened from radiated noise might create clicks in the audio. This can even happen if the noise isn't audible. If there is any stray signal present there is a short time when if the contact voltage is non-zero it gets transferred between paths as stray capacitance holds it momentarily as some random DC level. Ordinary PCB relays don't usually have any screening of the contacts and they are sticking up inside the plastic case like aerials.

For this relay, if you grounded the right most contact (NC) it would act as a shield.  Connecting the terminal to a low impedance output would achieve a similar result.  However connecting that terminal to a high impedance input would make things worse.   This is along the lines of what I was trying to get at at the end of Reply #47.  Except that was aimed at stopping the junk getting out.

The take home message is different relays, different contact choices and different circuits can make or break the performance of the build.

Test 1 : Powered the relay coil independantly, without the timer circuit, straight from an independant power supply+ momentary switch, without filtering
=>still got the tick+pop when switching

Test 2 : I applied strictly, on my breadboard, Jim's power supply filtering above's schematic for the switching&FX, fed from one power supply.
=> Good improvement, stability, buzzing is a lot better.

Those odd results point to something more going on.   If it were mine I'd probably try to unravel the mystery, however, these subtle issues never get solved on forums.  Perhaps a better plan is this:  In the case where you have separate supplies the idea is the circuits behave independently.  For any of the tests, even for the separate supply, add a 100uF to the power supply rail on the audio circuit and a 100uF cap to the power supply rail on the switching circuit.  The caps shouldn't be necessary but they should help suppress any secondary weird issues which mess with the results.  At least that way we can go forward.

[jackwithoneye]

Test 1 : Powered the relay coil independantly, without the timer circuit, straight from an independant power supply+ momentary switch, without filtering
=>still got the tick+pop when switching

Test 2 : I applied strictly, on my breadboard, Jim's power supply filtering above's schematic for the switching&FX, fed from one power supply.
=> Good improvement, stability, buzzing is a lot better.

Those results don't make sense as jim's circuit doesn't change the switching times.   It separates the power and test 1 already *is* separating the power.  In test 1 did you have a diode on the relay?

As a sanity test you should try replacing the relay entirely with a normal switch.  If that ticks/pops you have another issue which cannot be solved by playing with the relay circuit.

I've tested with a classic mechanical DPDT switch in place of the relay contacts (except the coils contacts of course), it's still getting the tick/pop, even louder than with the relay i think, and during Switching on and off with the mechanical DPDT


I've tested with the "audio" relay contact reverted (meaning that when the relay's coil is not powered- relay in idle position-, the effect is on, and when it's powered, the fx is bypassed). in this position the led is lit when fx is bypassed. Swithching On the fx (coil changing status from powered to unpowered) is still when the pop happens.

Considering these two tests, i tend to think that the relay (and timer circuit) is not generating the tick/pop by itself, when it's changing status, but it would be more likely the effects circuit INPUT changing from grounded(when bypassed) to Jack IN (when FX on) that would generate the pop. Am i wrong in my deduction?

I'm still thinking that some of my false or "make no sense" tests results analysis might be corrupted by the fact that i'm using a little bit "cheap" breadboard and non shielded wires.

test3 :
I had tried (maybe in a bad way, due to my misunderstanding? see below schematic) RG's BFC:
tried (in both direction) 100uF,47uF,10uF
It's actually slowing down the relay release (switching off), and making the "Zeeeew" light saber sound during this release, but doesn't affect the switching ON pop.

That connection works better for switching off.  When switching on the transistor shorts out the cap so the on-time is fast.  The turn-on speed is limited only by the transistor's ability to pass current (set by the base resistor).   So basically this circuit has extreme asymmetry in the turn on and turn off times.

To filter both on and off you need  different configuration.   Unfortunately most simple circuits like that one will end-up with the turn on being somewhat faster than the turn on.   You can for example add a resistor in between the relay+diode+added cap and the collector of the transistor.   The larger the resistor the slower the turn on time.   However, due to voltage drops, the resistor can't be too larger otherwise the relay won't turn on correctly.   The other way to make the turn-on time longer is to increase the added cap.  However that will also increase the turn-off time.   This is where the asymmetry in the on and off times becomes hard to work around.


If you still have the 10k base resistor, it might be worth trying a cap across the base and collector of the transistor again.

Tried differents low values resitors between collector and relay-diode without very good improvements, until the resistors gets too big (above 200R) and begin to mess up with the relay's coil activation.

Tha cap between transistor's base and collector doesn't really help, as it slows down the Switching off, but doesn't reduce switching on's pop. (tried different values, the only thing that changes it the switching of time length)